Working off @kaedonkers and @knapply's examples above, as well as @Bryan Shalloway's comment that these didn't work when specifying a base color for edges, I backed into the solution over here, with the key being the algorithm and bidirectional degree specifications within highlightNearest
library(igraph)
g <- graph("Zachary")
library(visNetwork)
vis_g <- toVisNetworkData(g)
visNetwork(vis_g$nodes, vis_g$edges) %>%
visIgraphLayout(layout = "layout_with_fr") %>%
visNodes(color = "blue") %>%
visEdges(color="blue") %>% # this causes prior solutions to not work
visOptions(highlightNearest = list(enabled = TRUE,
labelOnly = FALSE, hover = TRUE,
# but these two options fix it
algorithm="hierarchical",
degree=list(from=1, to=1)
),
nodesIdSelection = list(selected = 6))
By inspecting the source code of the npm package, I understood what exactly is that inputString parameter. Essentially, every schema method returns a Type instance which has two properties: inputString & isPrimitive. What I tried (and worked out) is storing a struct definition in a separate variable and pass it to the array method like
I believe that AGP 8.2.1 has some compatibility issues with Kotlin 1.9.20.
Also BaseVariant API was deprecated and basically removed in newer AGP versions.
To fix your build you should align your Kotlin version with your AGP version, all the info can be found via this link: AGP, D8, and R8 versions required for Kotlin versions
To further fix your build you should update your Gradle Wrapper to 8.6 distributionUrl=https\://services.gradle.org/distributions/gradle-8.6-all.zip
and also update your android/build.gradle to the right version:
To run Rust code in a Jupyter notebook, you need to install a Rust kernel for Jupyter. The most popular option is the evcxr_jupyter kernel, which allows you to write and execute Rust code in notebook cells, just like Python.
Here’s how you can set it up:
1. Install Rust (if you haven’t already)
If you don’t have Rust installed, you can install it with:
I'm using Vercel Hobby plan and worried about hitting the serverless function limit. My blog uses a server component that fetches posts via getBlogPosts() based on URL params. I'm also calling it in generateMetadata, [slug] page, and for sitemap/rss.
Even though I use generateStaticParams, it feels inefficient. Any way to reduce these invocations?
My thoughts so far:
Cache getBlogPosts() and reuse it across functions
Pre-generate sitemap/rss at build time
Move filtering to client side (if post count is small)
Use revalidate: false for static builds
Would love input from others who’ve optimized for similar c
I look for a script that when you run it, backup an actual system image of the Coral Dev Board. Because I didn't find it I develop one that copy the actual image of the Coral into an SD card.
When you want to restore the system to "factory reset" state, you produce an SD card with the system (mendel) image (enterprise-eagle-flashcard-20211117215217.zip) and with BalenaEtcher you bootable. Then you with the board off, change switches, put the factory reset SD and start. After that the board is in factory reset state.
The backup SD card works the same: after created by the script, you put it in SD reader, change switches, power on, and after that the Coral is in the same state that was backup created.
If somebody is interested in it I can send it or publish it.
Adding this for visibility and in case this is a widespread issue. I have a server that runs Google OAuth and at some point in the last 4 hours (since this was posted), this error has started occurring when a code exchange happens (OAuthClient.getToken(code)). The code was previously working - hoping more developers here can confirm this is an external bug or provide some solution.
You may check out this library that does all that, validate the South African ID and can also extract a lot of information from it. Like age, gender and citizenship: https://www.nuget.org/packages/SAIDValidator
I found this workaround: Before to run the compilation (cmake --build build ...), I edit the file build/CmakeCache.txt generated by cmake and replace the absolute Windows path C:/msys64/ucrt64/lib/libmingwthrd.a by the absolute path of the same file when you run the msys console : /ucrt64/lib/libmingwthrd.a
And the same for C:/msys64/ucrt64/lib/libgomp.dll.a --> /ucrt64/lib/libgomp.dll.a
I found an alternative solution that does not directly answer my question but still provides the same behavior I needed. Perhaps this is also me not fully understanding how authorization and custom domains works earlier.
I came across this video from a few years ago that explicitly mentioned making a second site in your Firebase project and linking a custom auth domain to it. This is what I did:
Inside my-project-staging Firebase project, I created a new site on the Hosting page and called it my-project-staging-auth. There is no code or custom app deployed to this project.
I added a custom domain to my-project-staging-auth called auth.staging.my-custom-domain.app
Added DNS config recommended by Firebase when going through the domain connection wizard.
Added https://auth.staging.my-custom-domain/__/auth/handlers to my Google client ID
Updated Firebase config on my app to use auth.staging.my-custom-domain for its authDomain property
This resolved the popup being handled by my main app. I'm still open to other alternatives if anyone has any, but this setup works well for my needs. Perhaps another way is to exclude certain routes in my React app so that Firebase handles them...
Unfortunately, you need to add them manually by clicking on the plus sign -> MCP Server -> Prompt/References. Maybe in a later version, this will be included so that it's aware of available prompts and resources and not just tools
I would like to add that using $CI_COMMIT_SHA (latest GIT commit) can be a good idea to track/detect a new application version. However, $CI_COMMIT_SHA is only reliable if every new/latest code change is a new commit that replaces the previous latest commit.
Developer might "reconstruct" a git branch history enough, that it will be effectively a new app version (based on actual code difference), but your pipeline or version detection logic will not detect that, because the latest commit SHA remained the same. Thus, your app may not notify the user of the change, for example, or some desired CI workflow won't happen because despite a change, CI_COMMIT_SHA has not changed. Or even if your app might still deploy with a new code version, not all deployment workflow steps will execute due to the latest commit SHA remaining the same.
try change in file vendor/magento/module-page-builder/view/adminhtml/web/template/page-builder.html tag iframe in sandbox attribute: if you have only allow-scripts add just allow-same-origin by patch. Works for me in 2.4.6-p10
#! /bin/sh
git rev-list --no-commit-header --format="%H %ae" HEAD |
while read commit_ID _email; do
if [ -z "$first_email" ]; then
first_email=$_email
fi
if [ "$first_email" = "$_email" ]; then
echo $commit_ID
else
exit
fi
done
We use the TEXT macro for the formatString for printf functions where code can be shared between windows UNICODE and UEFI ascii so the GCC warning is something I really want to bypass for UEFI ARM64. Trying to copy or something seems crazy and inefficient for this. GCC does not have <tchar.h> but I still need a way to make these printf's portable and seems reasonable to expect.
Your browser/OS might not support Object.hasOwn() if you haven't updated it in a while. It received wide support only about 2 years ago.
General Information
If you came here from a google search and you have the same error but it's unrelated to discord.js: Check your browser and OS.
For example, my job requires us to support all iOS devices that can run iOS 15+ (~4 years old at the time of writing). That's not unreasonable; but trying to run some external libraries crashed the app because Object.hasOwn() got support in iOS 15.4 (~2 years ago at the time of writing).=
I was never able to solve the problem. I had the company hardware people replace my computer with a new one and I was able to install VS2022 and the bug was gone. A few days later, after a security update to her computer, a third co-worker got the bug. She had the inspiration at one point to go over the ZScaler logs on her computer and found that VS was trying to communicate with the internet in a way that ZScaler blocked. I don't have the dirty details, but it seems the problem is a security issue.
First, you should have clarified where you are using such a block of statements (referred to as a "compound statement") and what you mean by "outside of" the block?
I ask, because generally, when you have such a block of statements, you use them as or inside of an SQL stored procedure or user-defined function (triggers, too, but that is a larger topic). Each of these have a means of returning a value or result set from the block.
Stored procedures can have parameters for output, "RETURN" a value or "data structure" (a data structure would be a concatenation of two or more columns into a single returned value that can be parsed and utilized by the caller), or pass back one or more "dynamic" result set(s).
User-defined functions can do much the same thing--the major difference being that calling a stored procedure is done as a separate SQL statement; whereas, calling a user-defined function is done inside of another SQL statement just as if it were a built-in function.
Tap on a clip to paste it in the text box.Tap on a clip to paste it in the teTap on a clip to paste it in the text box.Tap on a clip to paste it in the text box.xt Tap on a clip to paste it in tTap on a clip to paste it in the text box.Tap on a clip to paste it in the text box.he text box.Tap on a clip to pasI Ii of France and sleepwell te it in kitchen text box.
If sent within the Outlook user's account you need to send from their email address. Spoofing is not advised. If you want to route an ics file from a different email address you will need to set up smtp and send custom emails with an ics file attached. Alternatively, you can use Salepager to send Outlook calendar invites from a different email address.
Python docstrings can be written following several formats as the other posts showed. However the default Sphinx docstring format was not mentioned and is based on reStructuredText (reST)
The same error was happening to me, but I found out, the error is in the styled-components library which is not compatible with Expo 53, I uninstalled the library and the app worked normally.
In my case, I had custom build variants and multiple modules, and the production build varian was the selected one, just VERIFY THE BUILD VARIANT, in a normal project, the one that should be selected is the debug.
I pretty much think the problem is in the environment configuration where you have put in the secret key for token verification. Code is not issue as it is working in dev.
Once you have checked code, do check for CORS configuration and database connection.
Iterate the points on the line calculating the distance between then. This should be quite straighforward using UTM coordinates, but you probably can find some ready-to-go code in a lib.
Once you have a distance that's below a certain threshold, say 5 meters, you have your intersection.
Remember GPS in real world usage often max out the precision within 10m. Using multiple constelations 2-3 meters.
I am having the exactly some problem here.
I have two servers with Gluster 11 with 6 SSD DC disk as a bricks.
/dev/sdc1 1,8T 1,7T 120G 94% /data1
/dev/sdd1 1,8T 1,7T 123G 94% /data2
/dev/sdb1 1,8T 361G 1,4T 21% /data3
/dev/sdf1 1,8T 358G 1,4T 21% /data4
/dev/sdg1 1,8T 362G 1,4T 21% /data5
/dev/sdh1 1,8T 356G 1,4T 20% /data6
As you can see, the two first bricks are overused.
I have a gluster volume which use this bricks:
gluster vol status
Volume VMS is not started
Status of volume: stg-vms
Gluster process TCP Port RDMA Port Online Pid
------------------------------------------------------------------------------
Brick gluster1:/data1/vms 59976 0 Y 3896
Brick gluster2:/data1/vms 52513 0 Y 2565
Brick gluster1:/data2/vms 50314 0 Y 3978
Brick gluster2:/data2/vms 52867 0 Y 2652
Brick gluster1:/data3/vms 51747 0 Y 4071
Brick gluster2:/data3/vms 53994 0 Y 2741
Brick gluster1:/data4/vms 52358 0 Y 161845
Brick gluster2:/data4/vms 54324 0 Y 1570340
Brick gluster1:/data5/vms 54552 0 Y 161878
Brick gluster2:/data5/vms 54510 0 Y 1570373
Brick gluster1:/data6/vms 57117 0 Y 161911
Brick gluster2:/data6/vms 54696 0 Y 1570406
Self-heal Daemon on localhost N/A N/A Y 4106
Self-heal Daemon on gluster2 N/A N/A Y 2775
My question is:
If I do gluster vol VMS rebalance fix-layout and then gluster vol VMS rebalance start, there will be any impact on the VM disks, during this rebalance process?
When using pytest and verifying sys.exit() was called, or perhaps a SystemExit was directly raised, e.value.code contains the exit status only in a special case -- not in general.
BTW, we are discussing the "provisional" exit status, because, from https://docs.python.org/3/library/sys.html, trouble during Python cleanup could produce a different exit code.
The special case...
1. if sys.exit or SystemExit was called with a single int argument, like either of these statements sys.exit(2) raise SystemExit(2)
then e.value.code will match the exit status.
However, for these other cases, e.value.code WILL NOT match the exit status.
2. These statements, with no args sys.exit() raise SystemExit raise SystemExit()
will produce an exit status of 0, but e.value.code == 0 is false.
3. These statements, with a string arg, sys.exit('Trouble...') raise SystemExit('Trouble...')
will produce an exit status of 1, but e.value.code == 1 is false.
Any other usage of sys.exit or raise SystemExit is irregular, but for completeness, for these usages e.value.code WILL NOT match the exit status.
4. passing in a single obj that is neither int or str, or passing
multiple args to the SystemExit call, like sys.exit((4, 'Tribulation...')) raise SystemExit(4, 'Tribulation)
will produce an exit status of 1, but e.value.code == 1 is false.
So, if you needed to make an assertion about the exit status, one approach
would be to assert exitstatus(e.value) == ...
and implement python's SystemExit logic in an exitstatus function, like
import re
import sys
import pytest
def exitstatus(value):
"""Determine the exit status based on the SystemExit object's attrs.
More accurately, this is the *provisional* exit status, because
trouble during Python cleanup could produce a different exit code.
Example
with pytest.raises(SystemExit) as e:
sys.exit(2)
assert exitstatus(e.value) == 2
Example
with pytest.raises(SystemExit) as e:
raise SystemExit('Trouble...')
assert exitstatus(e.value) == 1
"""
if len(value.args) == 0:
return 0
elif len(value.args) == 1 and isinstance(value.args[0], int):
return value.args[0]
elif len(value.args) == 1 and isinstance(value.args[0], str):
return 1
else:
# Irregular usage of sys.exit() or SystemExit will produce
# exit status 1.
# If you want to trigger on any irregular usage so you can
# fix it, you could raise an exception here, to force your
# unit test to fail.
# Otherwise, well, the irregular usage does result in exit
# status 1, so this is faithful.
return 1
def test_sysexit_0():
with pytest.raises(SystemExit) as e:
sys.exit(0) # exit status should be 0
assert exitstatus(e.value) == 0
def test_sysexit_2():
with pytest.raises(SystemExit) as e:
sys.exit(2) # exit status should be 2
assert exitstatus(e.value) == 2
def test_sysexit_noargs():
with pytest.raises(SystemExit) as e:
sys.exit() # exit status should be 0
assert exitstatus(e.value) == 0
def test_sysexit_msg():
with pytest.raises(SystemExit,
match='cmd: bad usage') as e:
sys.exit('cmd: bad usage unrecognized option') # exit status should be 1
assert exitstatus(e.value) == 1
You discovered the root cause: your old Hibernate Validator version supports org.hibernate.validator.constraints.NotBlank but not the javax.validation.constraints.NotBlank properly.
Upgrading Hibernate Validator to 6.x fixes the problem, but you're stuck with 5.1.0 due to project constraints.
So your version of Hibernate Validator simply does not recognize or enforce the newer javax.validation.constraints.NotBlank annotation.
suggested solutions:
The cleanest approach without upgrading is to replace all usages of javax.validation.constraints.NotBlank with org.hibernate.validator.constraints.NotBlank in your code.
If you must stick to javax annotations for some reason, fallback to @NotNull + manual empty-string checks, or implement a custom validator and use it explicitly.
When your data contains quotes, CSV turns them into double quotes ("") and then wraps the whole field in quotes. So yes — CSV turns your input "foo014" into """foo014""" — that’s the correct and expected behavior.
"foo014"
Is interpreted as:
Quote → escape as ""
f
o
o
0
1
4
Quote → escape as ""
Then the entire field is wrapped in quotes, resulting in:
"""foo014"""
about your test:
Your actual output is correct — the test expectation is wrong.
To match Apache Commons CSV output, update your expected string to:
val expected = """
EPPN,FirstName,LastName,Email,ClassStanding,StudentID,Degree,College,Department,SuitableRole
"""foo014""",Foo,Bar,[email protected],,9200#8210,,,,ROLE_TENANT_ADVISOR
""".trimMargin()
p.MarkerClicked += async (s, args) =>
{
//I want to show custom content here
var p = (pin)sender;
await DisplayAlert("You clicked the pin", $"Pin was {p.Label}", "Cool");
};
It's more idiomatic, handles cleanup automatically and follows the principle of "context as the first parameter" which is a Go best practice
Manages multiple goroutines and their errors
Wait for all goroutines to finish
It's cleaner to read
1. Is there any fault in the new code?
It looks good to me
2. Compared to previous approach and this, which one is overall better and should be recommended over another? Second approach
3. Can I update the new code to make it better? Or any better graceful shutdown approach? This pattern is the idiomatic way to handle graceful shutdown, for most cases this is the best way to handle graceful shutdown
I need to do this also. I need the next 3 weeks of calendar entries, and a static url to retrieve that with. I can’t use dynamic url’s. Don’t really see what the point is of the calendar/events api if it doesn’t return data in calendarview format. Like OP says, it doesn’t return recurring events. But why would I want a set of calendar event data that doesn’t include certain entries just because they were created some time ago?
A querystring, also known as a query string, is a part of a URL that contains parameters and their values. It's used to send data from the client to the server, typically after the base URL and separated by a question mark (?).
Error on my part. Not only am I using ANSI 92, but I'm also restricting access to every table. and preventing visibility of the underlying database structure.
It appears that Microsoft.ACE.OLEDB.12.0 requires visibility...
I believe I figured this out. The Z_Value number I'm working with if I extend the digits out to 18, is actually 0.67450000000000044. Because there are the two 4's at the end, it is rounding up to 0.675.
try installing the qiskit 1.1.0 package. There are conflicts between qiskit versions that have not yet been resolved in version 2.0.2. Use the following command: "!pip install qiskit==1.1.0" . This version will probably solve your problem.
What's the range of possible values of these integers? If it's a relatively narrow range, a variation of Radix Sort (I call it Shawn Sort) has close to O(n) speed!
Say you're sorting millions of grades, which range from 1-100.
You make an array of 100 integers, with all elements initialized to 0.
Iterate over your dataset, incrementing the array element whose index matches the value (grade).
Traverse the resulting array, outputing the index (as a value) T times, where T is the array element's value.
Sorting a record (database table record, for example), works the same way, but you also store the identities or record numbers in the array, in which each element (which is a tuple) also contains a list of identities / record numbers.
Shawn Sort beats every other sort algorithm decisively, for integers, if the possible values is manageable relative to the size of the dataset.
I got the same issue this morning.
Found two ways to handle it:
1. (As a DBA) - added this separator in the query - ',' as ',', between columns. Then pasting was as usual.
2. (as a regular user) - closed all my Excel files and started it again. It fixed the issue.
You can obtain GlanceId by using LocalGlanceId.current, which can then be converted to an Int via GlanceAppWidgetManager.
val context = LocalContext.current
val glanceId: GlanceId = LocalGlanceId.current
val id: Int = GlanceAppWidgetManager(context).getAppWidgetId(glanceId)
Just, only downgrade your dart and flutter extension in vsc (unmark aut updates for these ext) and this solve the problem, sorry for my english. rgards.
I was using RevitAddinUtility.dll version 26.0.10.8 when I wrote up this question. Before I hit 'post' I realized I did not have the latest version of Revit 2026. I'm now using RevitAddinUtility.dll version 26.1.0.35, and the ManifestSettings save as expected.
The example code above may be useful to others, so I decided to go ahead and post this as a Q&A.
Going through each Rust version and checking the LLVM version it uses using rustc. A very accessible and convenient way to do this is using https://rust.godbolt.org/ which lets you switch between versions quickly.
I agree with XQ Hu. You might consider exploring the Bigtable Beam Connector (BigtableIO), which allows you to perform both batch and streaming operations on Bigtable data within a Dataflow pipeline. Additionally, you can create a custom container image using Docker to launch containerized SDK processes.
The first thing I notice is that your indexed range is not correct, you should index A2:C7. A quick solution I can think of is if you add an additional column that concatenates column A and column B, the formula would be =CONCATENATE(A2,B2), and then you drag it down up to your last row of data, then use. Let's say you put the CONCATENATE formula at column C, this will move the salary to column D. then you will have to use =INDEX(A2:D7,MATCH("Name",C2:C7,0),4)
I encountered the same issue with version 4.3*, but switching to version 4.29 resolved it. However, the local Maven build with version 4.3. runs without this issue, while I consistently face the same error in the Docker image unless version 4.29 is used.
In most browsers, using multiple favicons with media="(prefers-color-scheme: dark)" or even a fancy SVG that changes with CSS works great. But Safari? Not so much.
Here’s what’s going on:
Safari always picks the first favicon it sees in the HTML and ignores the rest, even if they have media queries.
It doesn’t support SVG favicons at all, so forget about using media queries inside SVG.
And even if you try switching the favicon with JavaScript, Safari often ignores those changes or just sticks with the original one — even across page loads, even if you try to bust the cache.
So… is there any way to make it work?
Option 1: Do it on the server (if you can)
If your site is server-rendered, you could try to serve a different favicon based on the user’s system theme. This involves detecting their preference before the page loads (which is tricky but possible in some setups using headers or early JavaScript). But honestly, it’s not super reliable and pretty advanced to pull off cleanly.
Option 2: JavaScript with a little hack
You can try forcing a favicon reload using JavaScript like this:
const dark = window.matchMedia('(prefers-color-scheme: dark)').matches;
const link = document.createElement('link');
link.rel = 'icon';
link.href = dark ? 'dark-icon.png?v=' + Date.now() : 'light-icon.png?v=' + Date.now();
document.head.appendChild(link);
This sometimes works — by adding a timestamp, you trick the browser into thinking it's a fresh file. But again, Safari is stubborn and may ignore this too.
Option 3: One neutral favicon
Honestly, the most reliable thing you can do is pick a single favicon that looks good in both light and dark mode. Something with a little border or contrast so it stands out no matter what.
TL;DR:
Safari just doesn’t play nice with dynamic favicons. You can try some workarounds with JavaScript, but they’re hit-or-miss. If it’s important for your brand, go with a high-contrast favicon that works in both themes.
HUBHIC:08371156:TCP: sck_complete_send() : Socket buffer is currently full (EWOULDBLOCK)
HUBHIC:08371156:TCP: sck_complete_send() : Not yet complete (bytes sent - 0 out of 555); retrying
HUBHIC:08371158:TCP: sck_complete_send() : Socket buffer is currently full (EWOULDBLOCK)
HUBHIC:08371158:TCP: sck_complete_send() : Not yet complete (bytes sent - 0 out of 555); retrying
HUBHIC:08371160:TCP: sck_complete_send() : Socket buffer is currently full (EWOULDBLOCK)
I have Visual Studio 2019 and a very large web application - and get this pop-up Extender error almost every time I open a solution and try to build it from a fresh start on the day. i.e. App opened Friday and compiled fine, Monday it fails - no changes...
The error always references remove and recreate the web.config
FIX is simple -
open web.config
press the space-bar
press backspace
The web.config is now in a changed state yet no changes were actually made
Click to run again and it works fine - no actual code changes were made.
I think the problem is due to the complexity of the application, it cannot process the previously complied web.config for some reason that is cached somewhere - changing it's state to changed even though technically it hasn't been changed causes Visual Studio to rebuild it again and it works.
